Important part of researches about data governance arereferring to the roles involved in Data Governance, a number of recommendationscan also be found in literature.Roles most frequently mentioned aredata stewards, data owners, and data committees. Data stewardssupport the business departments in thedesired use of data (when using standards like eCl@ss for classification of materials, for example). Also,data stewards are responsible for taking up and evaluating businessrequirements on and problems with data. Data stewards typically deal with datafrom a certain business department or division (Loshin, 2008). While datastewards represent an enterprise’s data management function, data owners belongto certain business departments or divisions. They specify the businessrequirements on the data and the data quality (Khatri and Brown, 2010).
Some researchers have criticized the term dataowner, as it might suggest that data “is owned” by a certain corporate functionor division, which would contradict the approach to view data as company assets(i.e. assets that are owned by the enterprise as a whole). Still, the term dataowner has established itself among practitioners.
The data owner role regardingsupplier master data, for example, is often allocated to the head of centralprocurement. A data committee is the central decision-making board in DataGovernance. It specifies the principles for using the data throughout theentire enterprise, and it matches the different interests and demands of thefunctional departments and business divisions (represented by the data owners)on the one side and the data management function (represented by the datastewards) on the other side (Khatri and Brown, 2010; Loshin, 2008).Regarding the linking of roles of workers withinthe company and decision areas, Data Governance is about assigning authorityand – resulting from this – responsibility. For example, the decision-makingauthority regarding the data architecture could be assigned to the data committee,whereas the executive power is assigned to the data steward.
When assigningroles to decision areas, the basic principle of congruence (in the sense oforganizational theory) is to be followed, according to which tasks,responsibility, and authority need to be congruent in order to ensure goaloriented action (Krüger 1994, pp. 47 f.). To do so, function diagrams are oftenused for modeling. A function diagram is a technique used in organizationaldesign by which tasks are linked with positions by means of so-called”authority codes” (Schulte-Zurhausen 2005, pp.
515 ff.).In order to design Data Governance for individualenterprises, a number of authors (Loshin 2008, pp. 33 ff.; Weber etal.
2009)propose to use the RACI notation. Besides identifying and describing eachof thethree elements of Data Governance (decision areas, roles, and authority),researchers are currently investigating the best possible combination of theseelements. Khatri and Brown (2010) speak of a “continuum” when assigningdecision making authority to central and/or decentral roles in enterprises.Weber et al. (2009) examined the use of contingency theory for the bestpossible design of Data Governance under consideration of enterprise specificexternal and internal parameters.